JP2020041290A - Optimal compaction determination construction system for concrete - Google Patents

Abstract

To enable anyone to easily and accurately determine the compaction of concrete when the concrete placed in a form is compacted without requiring complicated handling and to predict the condition with high accuracy.SOLUTION: An optimal compaction determination construction system for concrete, comprises: a video camera (imaging means) 5 for obtaining an image of concrete; an image feature amount calculation unit 24 which calculates, on the basis of the image acquired from the video camera 5, a numerical vector that is a feature amount relating to the degree of compaction of concrete by using machine learning; a learning unit 25 for deriving a predicted value of a numerical value representing the degree of compaction of concrete by using machine learning on the basis of the numerical vector calculated by the image feature amount calculation unit 24; and a compaction determination unit 27 that calculates a predicted value from the learning unit 25 to determine the compaction.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a concrete compaction determination execution system for determining an optimum compaction state when compacting fresh concrete.

  Normally, when compacting fresh concrete, it is required to remove air entrapped at the time of transportation and driving from the concrete, to make steel bars and buried objects adhere to the concrete well, and to make the concrete uniform and dense. Generally, compaction of concrete is performed by inserting a vibrator into concrete poured into a formwork and applying vibration. Also, in this case, the determination of sufficient compaction of the concrete, that is, the determination of compaction, is made by a veteran worker or technician by visually determining the concrete surface. Due to this vibration compaction, concrete is filled tightly to every corner of the formwork and reinforcing bars, and voids such as air bubbles inside and on the surface of concrete are eliminated, improving strength, watertightness and durability. Excellent concrete can be made.

  As a general guideline for adequate compaction of concrete, the concrete standard specification states that "one of the evidence that vibration compaction is sufficient is that a line of cement paste appears on the interface between the concrete and the dam. In addition, it can be seen from the fact that the volume of the concrete is no longer reduced, the surface of the concrete appears glossy, and the whole concrete appears to be melted uniformly. "

  Conventionally, a concrete compaction method for ensuring the required quality of cast concrete without being affected by the personal skills of workers has been proposed by Patent Document 1, and it has also been proposed to use camera photography and In addition, Patent Literature 2 and Patent Literature 3 propose a technique for determining compaction of cast concrete by a measurement technique using a distance sensor.

  In the concrete compaction method of Patent Document 1, a floating vibrator is installed at the bottom of a concrete form to be poured, and concrete is poured into the form so as to have a predetermined height. By levitation in the concrete while vibrating the vibrator according to the pressure, the cast concrete is compacted upward from the lower part.

  In this case, when the concrete to be poured into the formwork reaches a predetermined thickness, the prime mover is driven by a detection signal of the concrete pressure detected by the pressure sensor, whereby the vibrator vibrates at a predetermined frequency. Compacts the surrounding concrete and reduces the frictional resistance between the vibrator and the concrete. Also, since the bulk specific gravity of the vibrator is formed to be lighter than the specific gravity of the concrete, a known buoyancy is generated in the vibrator as the frictional resistance between the vibrator and the concrete decreases, and the vibrator vibrates to tighten the surrounding concrete. Ascends and rises upward. When the vibrator is positioned on the upper surface of the concrete, a detection signal having a concrete pressure of 0 is transmitted from the pressure sensor to the prime mover, the vibrator is stopped, and the compaction of the concrete is completed.

  Thus, in this concrete compaction method, by using a floating vibrator, the concrete compaction work can be automated and released from conventional heavy work, and the degree of concrete compaction can be reduced to personal skill. It does not rely on it, so it can improve the reliability of construction.

  In the technique for determining compaction of concrete disclosed in Patent Document 2, compaction work is performed by applying vibration with a vibrator inserted into the concrete. Here, the vibrator is temporarily stopped at a predetermined time interval set in advance, and the shape data of the concrete surface S is obtained by the shape data obtaining means (camera or the like). By the shape data analysis by the smoothness detecting means, the uneven shape of the surface is grasped, and the acquired smoothness of the surface is detected. Compaction degree determination means quantitatively determines the compaction degree of concrete based on the detected smoothness. As a result, when a predetermined degree of compaction is reached, it is determined that compaction is completed.

  Further, the technology for determining compaction of concrete disclosed in Patent Document 3 discloses a three-dimensional distance sensor that projects a light ray on an object to acquire three-dimensional distance data that is a set of point cloud data, A coordinate conversion unit that converts three-dimensional distance data into point cloud data based on three-dimensional orthogonal coordinates; a plane extraction unit that extracts a point cloud recognized as a plane from the converted point cloud data; A point group number calculation unit for calculating the number of point clouds of the point group constituting, and a determination unit for determining whether the calculated number of point groups is equal to or greater than a threshold value, the determination system comprising: When the number of point groups is equal to or greater than the threshold value, it is determined that the compaction of concrete has been completed.

JP-A-4-357274 JP-A-2017-025609 JP 2017-053084 A

  However, the conventional determination of compaction of concrete is performed by visual judgment of veteran workers and technicians on the concrete surface, and the criterion depends on the experience and skills of the workers and technicians, The number of people who can judge properly is limited, and on the other side, there may be individual differences in judgment results between workers and technicians, and furthermore, it is not possible to grasp the state of propagation of vibration inside concrete, There is a problem that the finish quality of concrete may vary.

  Moreover, although the concrete compaction method of Patent Document 1 has an advantage in that it is not affected by the personal skills of the worker, the vibrator itself floats, and thus the determination of concrete compaction is made by the vibrator in the concrete. In this case, there is a problem that the quality of the finished concrete may vary, because it is not possible to ascertain the state of propagation of the vibration inside the concrete.

  In the concrete compaction determination method disclosed in Patent Document 2, the vibrator is temporarily stopped at a predetermined time interval set in advance in order to acquire a determination image of cast concrete, and shape data acquisition means (camera or the like) is used. ), The work is interrupted because the shape data of the surface S of the concrete is acquired.

  Further, in the method for determining compaction of concrete disclosed in Patent Document 3, no photographing method such as a camera is used, and three-dimensional distance data is converted into point cloud data based on three-dimensional rectangular coordinates, and the point cloud data is converted. There is a problem that it takes time and effort for a human (designer) to perform a preliminary operation, such as designing a threshold value indicating the relationship between the concrete and the degree of compaction of concrete in advance.

  The present invention solves such a conventional problem, and an object of the present invention is to determine the compaction of concrete by performing the same operation as a visual judgment on a concrete surface by a veteran worker or a technician. An object of the present invention is to provide a concrete compaction determination and execution system in which the compaction determination is computerized and the concrete compaction determination eliminates individual differences in the experience and skills of workers and technicians.

  A second object of the present invention is to provide an optimum concrete compaction judging system capable of optimizing concrete compaction judging processing by a computer and performing high-quality compaction judgment.

  In order to achieve the above object, the present invention provides a vibrator that is inserted into cast concrete to apply vibration to the concrete and compacts the concrete, and a photographing unit that acquires an image of the compacted concrete And an image feature amount calculation unit that calculates a numerical sequence that is a feature amount related to the degree of compaction of the object using machine learning based on the image obtained from the shooting unit, and A learning unit that learns, using machine learning, a relationship between an image and the numerical sequence calculated by the image feature amount calculation unit, and derives a predicted value of a numerical value representing the degree of compaction of the concrete; and a learning unit. A compaction determination system for concrete, comprising: a compaction determination unit that performs a compaction determination by calculating a predicted value from the above. In the compaction determination system, the learning unit may derive a predicted value of a numerical value related to a surface state of the concrete as the predicted value.

  The present invention also obtains an image of the surface of the poured concrete by using a photographing means, and, based on the image obtained from the photographing means, by an image feature amount calculating unit, uses the machine learning to execute the concrete. A numerical value sequence, which is a feature amount related to the degree of compaction, is calculated, and the learning unit represents the degree of compaction of the object using machine learning based on the numerical value sequence calculated by the image feature amount calculating unit. The gist is an optimal compaction determination method for concrete in which a predicted value of a numerical value is derived, and a compaction determination unit calculates a predicted value from a learning unit to perform compaction determination.

  The present invention further provides a vibrator that is inserted into the cast concrete to vibrate the concrete to compact the concrete, a photographing means for acquiring an image of the compacted concrete, A vibrator driving means for performing a raising / lowering operation and a position changing operation of the vibrator; a photographing means driving means for performing a raising / lowering operation and a position changing operation of the photographing means with respect to the compacted concrete; Based on the image obtained from the means, an image feature amount calculation unit that calculates a numerical sequence that is a feature amount related to the degree of compaction of the object using machine learning, an image obtained from the imaging unit and the image The relationship with the numerical sequence calculated by the feature amount calculation unit is learned using machine learning, and the concrete compaction is performed. A learning unit that derives a predicted value of a numerical value representing the degree of compaction, a compaction determination unit that calculates a predicted value from the learning unit and performs compaction determination, and controls driving of the vibrator driving unit and the photographing unit driving unit. And an operation control unit.

  ADVANTAGE OF THE INVENTION According to the optimal compaction determination execution system of the concrete of this invention, when inserting a vibrator into the concrete poured into the formwork and applying a vibration to compact the concrete, the machine learning function of AI (Artificial Intelligence) is used. It is used to predict the degree of compaction of concrete and utilize it for compaction judgment.Therefore, in compaction judgment of concrete, the optimal compaction judgment of concrete excluding individual differences such as the experience and skills of workers and technicians and Construction can be performed. In addition, the compaction determination processing of the concrete by the computer is optimized, and the compaction determination of high quality can be performed.

It is a lineblock diagram showing the whole concrete compaction judging construction system concerning one embodiment of the present invention. FIG. 3 is a block diagram illustrating an example of a hardware configuration of a computer provided in the control device according to the embodiment. FIG. 3 is a functional block diagram illustrating a functional configuration of the control device illustrated in FIG. 2. It is a flowchart which shows the processing procedure of compaction determination and construction control by the control apparatus of the compaction determination construction system concerning the said embodiment. It is a process flowchart which shows an example of the operation | movement procedure by CNN employ | adopted by the learning operation | movement of said Embodiment. FIG. 7 is a diagram illustrating a state in which a labeled frame image acquired by an image acquiring unit changes with time by a compaction operation in the embodiment. It is a figure showing the quality of the compaction judgment result when the compaction test is performed to three kinds of concretes with different freshness in the above-mentioned embodiment.

  Next, an embodiment for carrying out the present invention will be described with reference to the drawings. FIG. 1 is a configuration diagram showing an entire concrete compaction determination execution system (hereinafter, simply referred to as “compaction determination execution system”) according to an embodiment of the present invention. As shown in FIG. 1, this compaction determination construction system 1 includes a form 3 for accommodating concrete 2, a vibrator 4 inserted into the concrete 2 in the form 3, and a vibrator 4 during compaction work. The camera includes a video camera 5 as a photographing means for photographing the surface of the concrete 2 near the insertion position, and a control device 6 connected to the video camera 5 to acquire a photographing signal from the video camera 5 and perform various data processing. Consisting of

  The vibrator 4 is attached to the tip (lower end) of a support rod 7 that suspends and supports the vibrator 4. The video camera 5 is attached to the tip (lower end) of a camera support rod 8 that suspends and supports the video camera 5. The support bar 7 and the camera support bar 8 are each movably connected to separate rails (not shown) and are moved by a drive inside the planar area of the formwork 3. A lift (not shown) is attached to the support rod 7, and the vibrator 4 is moved vertically with respect to the concrete 2 by moving the support rod 7 vertically by the lift. . Further, another camera elevating member (not shown) is also attached to the camera support rod 8, and the video camera 5 is moved up and down by the camera elevating member so that the video camera 5 is moved to the surface of the concrete 2. The height of the concrete 2 from the surface can be adjusted by moving it up and down. The video camera 5 is set so as to photograph the surface of the concrete 2 within a range of about 30 cm (centimeter) from the vibrator 4 by adjusting the position and the height.

  As described above, the video camera 5 and the control device 6 are in a connection relationship, and this connection relationship is such that the video camera 5 and the control device 6 are connected by a signal line indicated by reference numeral 9 in FIG. Alternatively, the video camera 5 and the control device 6 may be connected by a public communication network (such as the Internet) or a short-distance simple communication system (such as Bluetooth). Further, the video camera 5 is equipped with a distance measuring means such as a laser distance meter, and measures the distance between the concrete camera 2 and the video camera 5 to correspond to the difference in the scale of the acquired image. I have. Scale correspondence can also be performed by putting a scale in the acquired image. In the compaction determination execution system 1, the degree of compaction of the concrete 2 as an object is predicted by machine learning, and completion of compaction is determined. In addition, other than the video camera 5 described above, a digital camera, a 3D distance sensor, or any other means capable of capturing a two-dimensional array image can be used as the image capturing means.

  The machine learning used in the present embodiment is a pattern conversion for converting an image or an image feature into a numerical sequence by learning learning data which is a set of an image or an image feature and a known numerical sequence regarding the degree of compaction of concrete. This is a process of acquiring a function and predicting an unknown value using the pattern conversion function. In the present embodiment, a value at a future time point is predicted using teacher data as a numerical sequence and a pattern conversion function obtained from the teacher data. It should be noted that the numerical sequence is a sequence of numerical values obtained by various observation values of the phenomenon regarding the concrete 2, and is a sequence of numerical values obtained by observing the phenomenon based on a certain rule. In the present embodiment, an image to which information of compaction not completed (before) or compaction completed (just) is added as teacher data. As an example of machine learning, there is a Convolutional Neural Network (hereinafter, referred to as “CNN”). In the present embodiment, an example using deep learning (Deep Learning) in this neural network will be described.

  Other examples of machine learning include a recurrent neural network (RNN Recurrent Neural Network), a multi-layer perceptron (MLP Multilayer Perceptron), a support vector machine (SVM Support Vector Machine), or a regression corresponding to the SVM support regression. (SVR Support Vector Regression), decision tree learning, association rule learning, Bayesian network, and the like. The compaction determination execution system 1 according to the present embodiment uses any of these algorithms according to its usefulness. You may.

  An object to be predicted by the compaction determination execution system 1 according to the present embodiment includes an aspect of the surface of the concrete 2 as the compaction degree of the concrete 2 that has been cast.

  The compaction determination execution system 1 is configured so that an image signal acquired by the video camera 5 can be acquired by the control device 6 via the signal line 9 or the communication network. Further, the movement and the elevating operation of the vibrator 4 can be controlled by a control signal sent from the control device 6. Further, the moving and elevating operation of the video camera 5 can be controlled by a control signal sent from the control device 6.

  The control device 6 includes one or more computers. When the control device 6 includes a plurality of computers, each functional element of the control device 6 described later is realized by distributed processing. The type of each computer is not limited. For example, a stationary or portable personal computer (PC) may be used, a workstation may be used, or a mobile terminal such as a high-performance mobile phone (smartphone), a mobile phone, or a personal digital assistant (PDA) may be used. May be used. Alternatively, the control device 6 may be constructed by combining various types of computers. When a plurality of computers are used, these computers are connected via a communication network such as the Internet or an intranet.

  FIG. 2 is a block diagram illustrating an example of a hardware configuration of each computer 50 provided in the control device 6. The computer 50 includes a CPU (processor) 51 that is an arithmetic unit that executes an operating system, an application program for compaction determination, and the like, a ROM (read-only memory: read-only memory) that stores the application program, Alternatively, a main storage unit 52 composed of a RAM (random access memory: rewritable memory at any time) in which various data processed as needed in a data processing operation is stored, and an auxiliary storage unit composed of a hard disk, a flash memory, or the like 53, a communication control unit 54 including a network card or a wireless communication module, an input device 55 such as a keyboard and a mouse, and an output device 56 such as a display and a printer. Note that the hardware module to be mounted may vary depending on the type of the computer 50. For example, stationary PCs and workstations often include a keyboard and a mouse as input devices and output devices, and often include a monitor such as a display as an output device. In a smartphone, a touch panel functions as an input device and an output device. Often do.

  Each functional element of the control device 6, which will be described later, causes predetermined software to be read on the CPU 51 or the main storage unit 52, and operates the communication control unit 54, the input device 55, the output device 56, and the like under the control of the CPU 51. This is realized by reading and writing data in the main storage unit 52 or the auxiliary storage unit 53. Data and databases required for data processing are stored in the main storage unit 52 or the auxiliary storage unit 53.

  FIG. 3 is a functional block diagram showing a functional configuration of the control device 6 shown in FIG. The control device 6 is connected to the video camera 5 as a functional element, and is connected to the image acquisition unit 21 that inputs images continuously acquired by the video camera 5 and the main storage unit 52 or the auxiliary storage unit 53. A database 22 for storing teacher data and data necessary for various processes; a teacher data acquisition unit 23 connected to the database 22 for receiving teacher data from the database 22; an image acquisition unit 21 and a teacher data acquisition unit 23, an image feature value calculation unit 24 that calculates the feature value of the input image based on the data from the operation units 21 and 23, and a video camera that is connected to the image feature value calculation unit 24 and is a photographing unit. And learning the relationship between the image obtained from Step 5 and the numerical sequence calculated by the image feature amount calculation unit using machine learning. And a learning unit (prediction value derivation unit) 25 to derive the predicted value of the number representing the compaction degree of the concrete. The control device 6 is further connected to an operation unit 26 that performs an operation for compaction determination based on the processing results of the image feature amount calculation unit 24 and the learning unit 25, and is connected to the operation unit 26. A compaction determination unit 27 that determines whether the compaction state is compaction incomplete (before) or compaction complete (just), and a compaction determination unit 27 that is connected to the compaction determination unit 27. An output unit 28 that outputs the determination result to an output device 56 such as a monitor, and a compaction determination unit 27 that are connected to change the installation position of the vibrator 4 based on the determination result by the compaction determination unit 27, And an operation control unit 29 for performing a control operation of changing the installation position of the device. The image feature calculation unit 24 passes the received image as it is to the learning unit 25 or calculates the image feature and then transfers it to the learning unit 25 by the machine learning used in the learning unit 25.

  FIG. 4 is a flowchart showing a processing procedure of compaction determination and construction control by the control device 6 of the compaction determination construction system according to the present embodiment.

  First, when the processing of compaction determination execution control by the control device 6 is started in response to an instruction input or the like by a user, the vibrator 4 is lowered at a predetermined position (referred to as an insertion point A) in the mold 3, that is, Then, the descending operation is started, and the vibrator 4 is inserted into the concrete 2 (step S1). The vibrator 4 has a mechanism for raising and lowering the vibrator 4 with respect to the surface of the concrete 2, and is controlled by the control device 6.

  Next, the vibrator 4 is started (Step S2), and at the same time, the operation of the video camera 5 is started, and the surface of the concrete 2 that is being vibrated by the vibrator is photographed (Step S3). A mechanism for maintaining the distance between the video camera 5 and the surface of the concrete 2 and a distance measuring device (such as a laser distance meter) are mounted on the video camera 5 or the camera support rod 8, and are controlled by the control device 6. Reference symbol P1 in FIG. 1 represents a shooting range of the video camera 5 in the present embodiment. The photographing range P1 is, for example, a range approximately 30 cm (centimeter) from the vibrator 4.

  Next, the image data is acquired by the image acquisition unit 21 and the teacher data is acquired by the teacher data acquisition unit 22, and a two-dimensional array image is captured (step S4). In the capture processing of the two-dimensional array image, the image acquisition unit 21 receives images continuously acquired by the video camera 5 as image signals of a labeled frame image. The image acquisition unit 21 performs analog / digital conversion on the received image to generate image data, and sends the image data to the image feature amount calculation unit 24 as a two-dimensional array image. The teacher data acquisition unit 23 acquires the teacher data from the database 22 and sends the teacher data to the image feature value calculation unit 24 as a two-dimensional array image.

The operation of the compaction determination construction system according to the present embodiment includes a learning phase and a prediction (determination execution) phase. Therefore, the operation in the learning phase will be described first.
Learning phase In this learning phase, a set (learning data) of an image or an image feature amount and a known numerical sequence indicating the degree of compaction is stored in the database 22, and then the stored learning data is read out to execute machine learning. This is an operation mode in which the CNN is used to acquire a pattern conversion function. When the CNN is used for machine learning as in the present embodiment, the two-dimensional array image is input to the AI, that is, the CNN through the image feature amount calculation unit 24 (that is, through). Step S5). The CNN has learned in advance the correspondence between the frame image of the video input from the video camera 5 and the compaction state. Therefore, the input frame image is classified into a class having a high likelihood (likelihood). The image feature value calculation unit 24 calculates a numerical value vector (numerical value sequence) that is a feature value regarding the compaction degree of the concrete 2 as the target object by using machine learning based on the image data. For example, the image feature amount calculation unit 24 uses the image data and the teacher data to calculate the difference between the differential compaction amounts of the concrete 2 caused by the time difference in the future (for example, a predetermined time after the acquisition time of each image data). (Hereinafter, referred to as “differential compaction amount”) is performed by using supervised machine learning with a numerical value (objective variable) to be predicted by machine learning. Is a convolutional neural network.

  FIG. 5 is a processing flow chart showing an example of the operation procedure of the learning phase by the CNN. As shown in FIG. 5, the CNN executes a convolution process in the first convolution layer on the input image data 40 (Step S21), and then executes a pooling process in the first pooling layer (Step S22). Then, after that, the convolution processing in the second convolution layer is executed (step S23), the pooling processing in the second pooling layer is executed (step S24), and then the convolution processing in the third convolution layer 3 is executed. (Step S25), and thereafter, the convolution processing operation and the pooling processing operation are repeated, such that the processing of the fully connected layer A is executed (Step S26), and the processing of the fully connected layer B is executed (Step S27). In this way, the feature amount can be mechanically extracted from the image data. As an example in the case of using CNN, the image feature quantity calculation unit 24 repeats the processing in the convolutional layer and the pooling layer several times (for example, five times), and then performs the processing via the processing in the output layer. A predicted value of the difference compaction amount, which is a variable, is calculated. This predicted value indicates the degree of compaction of the concrete 2 in the future. At this time, the image feature quantity calculation unit 24 learns the relationship between the image and the numerical sequence related to the degree of compaction according to the CNN learning algorithm, and updates the pattern conversion function by updating the weight of the CNN combination. it can. Here, after constructing the pattern conversion function once, the image feature quantity calculation unit 24 may stop updating the pattern conversion function using the data of the actually measured values, and the pattern conversion function is constructed from the beginning. In this case, the function of updating the pattern conversion function may not be included. Further, the image feature quantity calculation unit 24 outputs the output of the output layer for the image data in the constructed state to the learning unit 25 as a numeric vector (for example, a numeric vector of 256 columns) as a feature quantity. I do.

Next, the prediction (determination execution) phase will be described.
Prediction (judgment execution) phase This prediction (judgment execution) phase is an operation mode in which a numerical sequence is predicted by using the CNN obtained by the machine learning for the pattern conversion function obtained in the learning phase. FIG. 6 shows an operation example of the prediction (determination execution) phase. FIG. 6 is a diagram showing how the labeled frame image obtained by the image obtaining unit 21 changes with time due to the compaction operation. As shown in FIG. 6, immediately after the concrete 2 is cast (after 0 seconds), compaction has not yet progressed, and the compaction determination result is a result of compaction not completed (before). In a state four seconds after the concrete 2 has been poured, compaction has not yet sufficiently progressed, and the compaction determination result remains compaction incomplete (before). In a state 8 seconds after the concrete 2 is cast, the compaction has sufficiently proceeded, and the compaction determination result is a result of compaction completion (just). In a state 16 seconds after the concrete 2 is cast, the compaction has progressed more sufficiently, and the compaction determination result indicates that compaction has been completed (just).

  The learning unit 25 derives a predicted value of the difference compaction amount as a numerical value representing the compaction degree of the concrete 2 based on the numerical value vector calculated by the image feature amount calculating unit 24. An example of the operation of deriving the predicted value will be described. The learning unit 25 executes machine learning based on a numerical vector obtained by combining a numerical vector and time-series environmental measurement data to obtain a difference as an objective variable. The transition of the predicted value of the compaction amount is calculated. At that time, the learning unit 25 updates the pattern conversion function such as a filter parameter by executing machine learning based on the teacher data. After constructing the pattern conversion function, the learning unit 25 may stop updating the pattern conversion function using the data of the actually measured values. The function update function need not be included. Then, the learning unit 25 outputs to the compaction determination unit 27 data on the transition of the predicted value of the difference compaction amount by machine learning in a state where the pattern conversion function has been constructed.

  The compaction determining unit 27 determines whether the compaction degree of the concrete 2 is not compacted or compacted based on the transition data of the predicted value of the differential compaction amount (step S6), If compaction has not been completed, a "before" signal is output, and the process returns to the shooting operation by the video camera 5 in step S3. On the other hand, if compaction is completed, a "just" signal is output. Further, the compaction determination unit 27 sends the “before” signal or the “just” signal to the operation control unit 29.

  The operation control unit 29 generates a control signal for controlling the operations of the vibrator 4 and the video camera 5 based on the compaction determination result of the concrete 2 output by the compaction determination unit 27. When the compaction determination unit 27 outputs the “just” signal, the operation control unit 29 starts the operation of winding the vibrator (step S7), and after performing this operation, stops the vibrator 4 (step S7). S8).

  Next, the CPU 50 of the control device checks whether or not the compaction of the entire area of the concrete 2 has been completed (step S9). The movement is instructed to the operation control unit 29 (step S10). The operation control unit 29 moves the support rod 7 and the camera support rod 8 along the corresponding rails, and changes the compaction position of the vibrator 4 from the insertion point A to the insertion point B as shown in FIG. Further, the shooting area of the video camera 5 is changed from the shooting range P1 to the shooting range P2. Then, the operations of steps S1 to S9 are performed by the processing operation of the control device 6.

Example A compaction test was performed on three types of concrete having different freshness. In this compaction test, 50 compaction test videos obtained by shooting video of 30 frames per second by the video camera 5 were obtained.

  Then, from the compaction test video obtained by capturing 30 frames of video per second, an average value of an appropriate time (specified by a frame number) of compaction completion (just) determined by a plurality of skilled operators is obtained, This average value was used as teacher data at the boundary between compaction not completed (before) and compaction completed (just) in the compaction test of the concrete. In addition, all the frame images acquired by the image acquisition unit 21 were given the correct label of "before" or "just", and were subjected to a "compaction determination" and an "evaluation of compaction determination" tests. Further, in consideration of the difference in the compaction state of the concrete depending on the position of the vibrator during the compaction test, the frame image was divided into two left and right images, and the appropriate time for each was determined.

  FIG. 7 is a diagram illustrating the quality of the compaction determination result when the compaction test is performed on the three types of concrete 2 having different freshness as described above. In FIG. 7, Test 01, Test 02, and Test 03 represent compaction tests for three types of concrete 2 having different freshness. In each test, the upper row represents the boundary of “before” or “just” of the compaction determination in the teacher data, and the lower row represents the quality of the determination result by the compaction determination execution system according to the present embodiment. Regarding the quality of the lower determination result in FIG. 7, the determination result “correct” indicates that the determination result was good in both “before” and “just” of the compaction determination. On the other hand, the determination result “incorrect” indicates that the determination result was incorrect in both “before” and “just” of the compaction determination. As shown in FIG. 7, in the test 01, the judgment result (lower part) by the judgment and construction system with respect to the teacher data (upper part) is not so far from the boundary part of “before” or “just” and the boundary part. Since a determination error (black portion) appears in the range, a determination result matching the outline teacher data is obtained.

  In test 02, the judgment result (lower row) of the teacher data (upper row) by the determination / construction system is not only at the above “before” or “just” boundary, but also in the “before” direction from the boundary (about 550 frame positions) Also, in the "just" direction (positions ahead of about 1600 frames), a determination error (black portion) appears, so that a considerably inaccurate determination result is obtained.

  In the test 03, the judgment result (lower part) by the judgment and construction system with respect to the teacher data (upper part) shows an erroneous judgment (black part) only in a portion close to the boundary of the above "before" or "just". As a result, a determination result that is very consistent with the teacher data is obtained.

  In this compaction test, only an image obtained by resizing the right image of the frame image to 64 × 64 pixels was used.

  According to the present invention, in the concrete compaction determination, the degree of concrete compaction is predicted using the machine learning function of the AI and used for the compaction determination. This makes it possible to judge the optimum compaction of concrete and to perform construction without differences between individuals such as experience and skills, which is useful.

1 Compaction judgment construction system 2 Concrete 3 Formwork 4 Vibrator 5 Video camera (photographing means)
6 Control device 7 Support rod 8 Camera support rod 9 Signal line 21 Image acquisition unit 22 Database 23 Teacher data acquisition unit 24 Image feature quantity calculation unit 25 Learning unit (predicted value derivation unit)
26 operation unit 27 compaction determination unit 28 output unit 29 operation control unit

Claims (4)

A vibrator that is inserted into the cast concrete to give vibration to the concrete and compact the concrete,
Photographing means for acquiring an image of the concrete to be compacted;
Based on the image obtained from the photographing means, based on the image, an image feature amount calculation unit that calculates a numerical sequence that is a feature amount related to the degree of compaction of the object using machine learning,
The relationship between the image obtained from the photographing means and the numerical value sequence calculated by the image feature amount calculation unit is learned using machine learning, and a predicted value of a numerical value representing the degree of compaction of the concrete is derived. Learning department
A compaction determination unit that performs a compaction determination by calculating a predicted value from the learning unit;
Optimal compaction determination system for concrete equipped with The learning unit derives a predicted value of a numerical value related to the surface state of the concrete as the predicted value,
The optimal compaction determination system for concrete according to claim 1. Using the photographing means, obtain an image of the surface of the cast concrete,
By the image feature amount calculation unit, based on the image obtained from the photographing means, using a machine learning to calculate a numerical sequence that is a feature amount related to the compaction degree of the concrete,
The learning unit learns, using machine learning, the relationship between the image obtained from the photographing unit and the numerical sequence calculated by the image feature amount calculation unit, and obtains a numerical value representing the degree of compaction of the concrete. Derive the predicted value,
The compaction determination unit calculates the predicted value from the learning unit and performs compaction determination.
A method for judging the optimum compaction of concrete, characterized in that: A vibrator that is inserted into the cast concrete to give vibration to the concrete and compact the concrete,
Photographing means for acquiring an image of the concrete to be compacted;
Vibrator driving means for performing a raising and lowering operation and a position changing operation of the vibrator with respect to the poured concrete,
For the compacted concrete, a photographing means driving means for performing a raising and lowering operation and a position changing operation of the photographing means,
Based on the image obtained from the photographing means, based on the image, an image feature amount calculation unit that calculates a numerical sequence that is a feature amount related to the degree of compaction of the object using machine learning,
The relationship between the image obtained from the photographing means and the numerical value sequence calculated by the image feature amount calculation unit is learned using machine learning, and a predicted value of a numerical value representing the degree of compaction of the concrete is derived. Learning department
A compaction determination unit that performs a compaction determination by calculating a predicted value from the learning unit;
An operation control unit that drives and controls the vibrator driving unit and the photographing unit driving unit;
Optimum compaction judging construction system with concrete.